RE: agentive -wI'

[eric.andeen@autodesk.com]

From: <cheesbro@rpa.net>
> HoHwI'wI' - "my killer".  Did I hire him to kill someone else, or did someone
> else hire him to kill me?
>
> muHoHwI' - "one that kills me".  Someone hired him to kill me.

SuStel:

> This sort of thing only makes semantic sense when the subject of the 
> verb-that-isn't is third person.  The following example words would 
> seem to be nonsense:

> choHoHwI'
>  qaHoHwI'
>  juHoHwI'
>  jIHoHwI'
>  maHoHwI'

Just to play Fek'lar's advocate, why are these nonsense? <choHoHwI'> - you who kill me. An example: <qatlh chojoy'nIStaH choHoHwI' mIgh? HIHoH neH jay'!>. To me, this is not really much stranger than <muHoHwI'>.

My feeling for the <-wI'> suffix is that the person (or object) described by the verb has the verb as his defining characteristic, at least in whatever context it gets used. A <HoHwI'> is someone who kills - that's what he does. 

The moment you start adding verb prefixes, you get into specifying who is being killed and maybe even who is doing the killing - you start talking about specific situations rather than general description. That just doesn't fit. It's time for <-bogh>.

Note that I do *not* have a problem with adding suffixes to verbs with <-wI'>, within reason. A <QoylaHbe'wI'>, to use Qov's example, is unable to hear - that's his defining characteristic in this context.

> The comparatively large number of these, combined with a restriction 
> on prefixes and /-ghach/ given to us by Okrand, are sufficient to 
> convince me that prefixes shouldn't be used on verbs with /-wI'/, 
> even if you could glean something sensical out of it.

> Clarity of */muHoHwI'/ over /HoHwI'wI'/ is nice and all, but often 
> the ambiguous things are the ones that win.

DaH wa' DoS wIqIp.

pagh